Using Polymer Electrolyte Membrane Fuel Cells in a Hybrid Surface Ship Propulsion Plant to Increase Fuel Efficiency by Douglas M. Kroll B.S., Electrical Engineering, Virginia Tech, 2001 Submitted to the Department of Mechanical Engineering and Engineering Systems Division in Partial Fulfillment of the Requirements for the Degrees of Naval Engineer and Master of Science in Engineering and Management ARCHIVES MASSACHUSETTIS INSTITUTE OF TECHNOLOGY at the Massachusetts Institute of Technology June 2010 SEP 0 1 2010 LIBRARIE S @ 2010 Douglas M. Kroll. All rights reserved. The author hereby grants to MIT permission to reproduce and to distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium now known or hereafter created. Signature of Author................................................................. Departmn of Mecanical Engineering and Systems esign and Manageme t Program /9 x Certified by.....................................................................f............ May 7,2010 .. Mark S.Welsh Professor of the PracticNaYal Construction and En' eti Acce pted by.................................................................................. ....------ u ring isor aPat Hale Director, System Design and Management Fellows Program Fpwineering Systems Division Accepted by..................................................................................... David Hardt Chairman, Department Committee on Graduate Studies Department of Mechanical Engineering Using Polymer Electrolyte Membrane Fuel Cells in a Hybrid Surface Ship Propulsion Plant to Increase Fuel Efficiency by Douglas M. Kroll Submitted to the Department of Mechanical Engineering and Engineering Systems Division on May 7, 2010 in Partial Fulfillment of the Requirements for the Degrees of Naval Engineer And Master of Science in Engineering and Management Abstract An increasingly mobile US Navy surface fleet and oil price uncertainty contrast with the Navy's desire to lower the amount of money spent purchasing fuel. Operational restrictions limiting fuel use are temporary and cannot be dependably relied upon. Long term technical research toward improving fuel efficiency isongoing and includes advanced gas turbines and integrated electric propulsion plants, but these will not be implemented fleet wide in the near future. The focus of this research isto determine if a hybrid fuel cell and gas turbine propulsion plant outweigh the potential ship design disadvantages of physically implementing the system. Based on the potential fuel savings available, the impact on surface ship architecture will be determined by modeling the hybrid fuel cell powered ship and conducting a side by side comparison to one traditionally powered. Another concern that this solution addresses isthe trend in the commercial shipping industry of designing more cleanly running propulsion plants. Thesis Supervisor: Mark S.Welsh Title: Professor of the Practice, Naval Construction and Engineering Thesis Supervisor: Pat Hale Title: Director, System Design and Management Fellows Program Acknowledgements The author would like to thank the following for their assistance: e e * * The instructors at the MIT Naval Construction and Engineering Program, Captain Patrick Keenan, Captain Mark Welsh, Commander Trent Gooding, and Commander Joe Harbour. Thank you for so generously putting your careers on hold to come here and broaden our views on both the technical and practical aspects of US Navy Acquisition. The fuel cell team at the Navy Surface Warfare Center in Philadelphia for sparking my interest in this field. Pat Hale for the patience and knowledge to continuously improve the System Design and Management Program. Captain Mark Welsh whose subtle and not so subtle motivational techniques have taught me to always dig deeper when difficult questions arise. Table of Contents --------. .............................. 3 Abstract ..............................................................................................-------...... ----------..............---- 4 Acknow ledgem ents.............................................................................................-----.... .. ----....... ----------.............. 5 Table of Contents ........................................................................................................... ---. ----------------................... 6 List of Figures .............................................................................................. 1. . . 9 -..............-----..... Introduction..................................................................................................... ...9 Problem Statem ent ...........................................................................................................-... 1.1 11 ........... Establishing Context................................................................................................. 1.2 12 Com plications of Implem enting Cost Reductions .................................................................. 1.3 .......... ....... ----.... 13 Research Proposal ......................................................................................... 1.4 2. .- -.... -14 Fuel Cell Discussion ............................................................................................................. 14 Electrochem ical Reaction Description .................................................................................... 2.1 Inherent Fuel Cell Losses.......................................................................................................-....16 2.2 --...........--------... 21 Types of Fuel Cells ..........................................................................................2.3 21 Solid Oxide Fuel Cells (SO FC)........................................................................................... 2.3.1 22 Polym er Electrolyte M em brane Fuel Cells (PEM FC) ...................................................... 2.3.2 24 Phosphoric Acid Fuel Cells (PAFC)................................................................................. 2.3.3 ..---- 25 Fuel Cell Selection ........................................................................................................----. 2.4 25 Durability Concerns ..................................................................................................................... 2.5 27 3. Propulsion Plant M odel ....................................................................................................................... 27 General Assum ptions .................................................................................................................. 3.1 28 Gas Turbine Inputs .............................................................................................................. 3.1.1 31 PEM FC Inputs ...................................................................................................................... 3.1.2 33 DDG-51 Flight 1 Inputs .................................................................................................... 3.1.3 34 M odel Algorithm and code....................................................................................................... 3.2 37 4. Optim ization Results ........................................................................................................................... .... --.37 M odel Output...............................................................................................................4.1 45 Sensitivity Analysis ...................................................................................................................... 4.2 45 Varying the Speed Profile ............................................................................................... 4.2.1 49 Varying PEM FC efficiency ................................................................................................ 4.2.2 50 5. Hybrid Propulsion Plant integration................................................................................................ 57 6. Sum m ary ............................................................................................................................................. Bibliography.........................................................................................................................................59 60 Appendix: Propulsion Plant M odel Code ............................................................................................... 7. List of Figures Figure 1: Fuel Cell Triple Phase Boundary (O'Hare, 2009) ...................................................................... 14 Figure 2: Fuel Cell Flow Paths (O'Hare 2009).......................................................................................... 16 Figure 3: Fuel Cell Activation Losses...................................................................................................... 17 Figure 4: Fuel Cell Ohmic Losses ................................................................................................................. 18 Figure 5: Fuel Cell Concentration Losses............................................................................................... 19 Figure 6: Fuel Cell Voltage Current Relationship with Irreversible Losses............................................. 20 Figure 7: Gas Turbine Specific Fuel Consumption (Woud, 2002)........................................................... 28 Figure 8: Total Fuel Consumption of 4 Gas Turbines ............................................................................. 29 Figure 9: Fuel Consumption Differential Based on Operating Method ................................................. 30 Figure 10: Gas Turbine Fuel Consumption at 3 Loading Conditions ...................................................... 31 Figure 11: Fuel Cell Voltage Vs. Current Density.................................................................................... 32 Figure 12: DDG 51 Flight 1 Powering Curve .......................................................................................... 33 Figure 13: Propulsion Plant Optimization Algorithm ............................................................................ 34 Figure 14: Total Fuel Consumption up to 10 MW ................................................................................. 38 Figure 15: Total Fuel Consumption up to 100 MW ................................................................................. 39 Figure 16: Fuel Consumption Differential for a Single Inefficiency Region............................................. 40 Figure 17: Fuel Consumption Difference Family of Curves ................................................................... 41 Figure 18: Total Fuel Consumption for a DDG-51 Speed Profile ............................................................. 42 Figure 19: Fuel Cell Operating Scheme up to 10 MW ............................................................................. 43 Figure 20: Fuel Cell Operating Scheme up to 100 MW .......................................................................... 44 Figure 21: Decreasing Operation at All Inefficient Speeds......................................................................46 Figure 22: Steaming Profile Sensitivity Analysis for all Low Speeds...................................................... 47 Figure 23: Decreasing Operation at a Single Inefficient Speed ............................................................... 48 Figure 24: Steaming Profile Sensitivity Analysis Shifting Away from the 15 Knot Speed ....................... 48 Figure 25: Consumption Sensitivity based on Fuel Cell SFC.................................................................... 50 Figure 26: DDG-51 Machinery Layout ................................................................................................... 52 Figure 27: Hybrid Propulsion Plant..............................................................................................................55 List of Tables Table 1: Fuel Cell Characteristics Comparison (O'Hare 2009)................................................................ 25 Table 2: Fuel Cell Operating Parameters............................................................................................... 32 Table 3: DDG-51 Flight 1 Mission Speed Profile...................................................................................... 34 Table 4: Operating Scheme for a 20 Fuel Cell Propulsion Plant............................................................. 44 Table 5: Summary of Speed Profile Sensitivity Analysis......................................................................... 49 Table 6: ASSET* 5.3 DDG-51 Design Summary...................................................................................... 53 Table 7: Ship and Equipment Modifications .......................................................................................... 54 Table 8: Hybrid Ship Principle Characteristics........................................................................................ 56 Table 9: Principle Characteristics Comparison Table ............................................................................. 57 Page Intentionally Left Blank 1. Introduction Over the past decade the US Navy has expanded its mission from the traditionally defined role of maintaining a sea line communication presence to assisting in the global war on terror and other forward deployed operations. Over time this shift will require that surface ships and submarines are able to quickly relocate to a more loosely defined set of locations. One way this transition can be achieved is by adding to the total number of ship hulls in service as well as requiring that each individual ship transit for a larger portion of its steaming profile. Due to the extensive costs associated with designing, building, and manning new ship classes, increasing the number of in service hulls has been occurring slowly. Additionally, the unstable cost of fuel has made operating them with this level of flexibility more expensive. 1.1 Problem Statement Two specific problems arise as the Navy requires a naval ship possess more mobility. The first isthe operational cost associated with the higher overall fuel use. This cost isdetermined primarily by the price of oil which isdriven by market supply and demand. The second isan indirect technical cost which is realized in maintaining an adequate hull form and displacement in spite of the larger required fuel storage volume. As aship's displacement isincreased, higher construction costs are immediately realized due to the larger quantity of fabrication material. Additionally, as displacement and submerged surface area becomes larger, an increase in the hull's frictional and formal drag raise the ships overall hydrodynamic resistance. This higher resistance must be overcome and can be represented as a decrease in propulsive efficiency, which then requires an even larger fuel storage volume. Several viable options are available to deal with the trend of increasing required surface ship mobility. The first isto install a more robust military support infrastructure to ease forward deployed refueling. This would allow ships to travel less distance between refueling and the depots could obtain overseas oil without having to pay transportation costs. The disadvantages of this solution are that the fuel price is still driven by external factors, and processing the fuel would have to be paid for or conducted at the remote site. Another method available would be to focus on technological advances which increase propulsive and electrical efficiency. This could be accomplished by advanced prime movers and motors such as high temperature superconducting motors, inter-cooled recuperated gas turbines, and an all electric based ship architecture. Although there ispromise in reducing power losses based on ongoing research, these are not achievable in the near term and become expensive to implement once the technology does become mature. The third potential solution for new ship designs isto use diesel generators in the place of gas turbines which are more fuel efficient in all loading conditions. One disadvantage isthat diesel generator volume and displacement far exceed that of gas turbines. This means any fuel savings advantage which could lead to smaller ship displacement would be diminished by the larger weight requirements of the more bulky diesel powered prime movers. Another disadvantage of this solution isthat with current diesel engine technology, a significantly higher amount of harmful effluent fumes are released compared to gas turbines. Afinal solution available isto use an alternate propulsion source such as electromechanical fuel cells to either completely power or supplement currently installed gas turbines in meeting the ship's propulsion requirement. This method isbeneficial for several reasons. These types of electricity sources, in addition to being more efficient than gas turbines, operate with significantly fewer moving parts. Less parts leads to quieter and more reliable operation. These characteristics allow for efficient power but also assist the ship remain stealthy and ensure propulsion remains continuous. Both of these benefits contribute to the ship's overall survivability. Although there are benefits, a problem arises while attempting to implement this solution within present day propulsion plant configurations. An efficient conversion process to couple dissimilar sources of prime mover energy and still allow for engine room flexibility isnot yet available and electric motors in use do not possess the capability of simultaneously interacting with both an electric and mechanical power source. Because of this, complex mechanical linkages would be required to achieve the desired configuration, which drives up fabrication and maintenance costs. Another disadvantage to fuel cells isobtaining the fuel source. Typical fuel cells are powered by hydrogen which must be either loaded directly from off hull sources or generated on board. As there is no hydrogen infrastructure in place it isassumed onboard reforming of gas turbine fuel would be required for any shipboard use of fuel cells in the near future. 1.2 Establishing Context To lower its operational budget, the Navy has placed an emphasis on decreasing overall fuel use. However, this philosophy contradicts with missions which tend to increase a given hull's cumulative transit time. Due to these differing pressures, more efficient ship designs seeking to raise propulsion plant efficiency while at the same time meeting battle group speeds are highly desirable. Inaddition to engineering improvements, a larger focus isbeing placed on lowering foreign oil dependency. This process once achieved will benefit the US Department of Defense by creating more stable fuel pricing. Until the US implements steps to make this occur however, the Navy will go through periods where prices rise uncontrollably and operational tradeoffs will have to be made to remain financially responsible. Steps toward this goal will take time making it a long term process. Any factors that can quickly mitigate excessive fuel consumption are therefore urgently needed. Another consideration which makes addressing this problem urgent isthat commercial shipping regulations are becoming more restrictive toward ensuring cleaner running ships are designed and operated. Many recently implemented laws are aimed at reducing or eliminating completely the release sulfur, carbon monoxide, and other harmful propulsion plant effluents. Any long term shifts in US Navy propulsion plant acquisition will most likely need to address environmental concerns to stay in step with industry standards. 1.3 Complications of Implementing Cost Reductions The problem of Department of Defense cost overruns isnot new and has in the past been engaged at both the acquisition and operational level. To lower fuel costs, solutions such as manually restricting speed and monitoring fuel use have been proposed and in some cases implemented. These measures however do not address the underlying problem of ships that are designed for high speed operation but have poor fuel efficiency at low speeds. These fixes are also short term solutions which tend to be responsive in nature and will not continue to provide benefit as operational constraints change. The difficulty in implementing tangible solutions isthat decision making within the design process isvery political. Current spending habits are engrained. A high level of ship performance and component level technological advance has also been deemed acceptable in new construction projects. This leads to ship operational requirements which drive the ship to perform a broad range of functions well. For example, having a large number of ship classes required to meet aircraft carrier battle group speed regardless of ship type, displacement or primary mission. This has led to the higher fuel consumption gas turbines being more prevalently selected within ship designs due to their high performance level. 1.4 Research Proposal The focus of this research isto determine if an optimized propulsion plant utilizing the fuel savings of a hybrid fuel cell and gas turbine propulsion engine outweigh the potential ship design disadvantages of physically implementing the system. My intention isto categorize three possible fuel cell types to highlight their advantages and disadvantages. Inherently driven inefficiencies as well as durability issues will also be discussed. This will allow the development of an accurate voltage-current relationship over the possible fuel cell operating power levels. Using this background information, one fuel cell type will be selected and developed further for a specific shipboard application. Once the parameters and assumptions are determined for the fuel cell of interest, an optimizing program will be generated using MATLAB*. This program will analyze afamily of notional hybrid propulsion plants across a given spectrum of operational profiles and compare the potential fuel consumption to currently seen levels. Finally, I intend to implement the optimized solution for hybrid fuel cell capacity at a total ship design level. Based on the potential fuel savings available I will qualitatively determine the impact on surface ship architecture by modeling the hybrid fuel cell powered ship and conducting a side by side comparison using the Advanced Surface Ship Evaluation Tool (ASSET*) version 5.3. 2. Fuel Cell Discussion Describing the general electric chemical reaction isthe first step toward deciding if fuel cells as an alternate energy source can be feasibly implemented within a naval engineering application. Once the basis for operation has been established, the inherent electric losses which determine the voltage to current relationship will be identified and described. Next, the advantages and disadvantages of three prominent fuel cell types will be highlighted, with the most applicable selected for implementation analysis. Finally, several durability concerns of the desired fuel cell type, and how they may hinder implementation will be identified. 2.1 Electrochemical Reaction Description Fuel cells generate electricity by using fuel in gaseous form to create atriple phase boundary, shown in Figure 1,which consists of fuel, electrolyte, and electrode. When this occurs, the heterogeneous oxidation process removes electrons from the charge carrier. The fuel after having been activated, transfers free electrons to the electrode across the catalyst while the charge carriers are driven into and across the electrolyte. Catalytic Electrode Particles Gas Pores TPB's Electrolyte Figure 1: Fuel Cell Triple Phase Boundary (O'Hare, 2009) The oxidation reaction can occur with multiple fuel source and electrolyte material combinations which are capable of stripping electrons from the base compound. Atypical oxidation reaction with hydrogen as the fuel is shown in Equation 1. H2 -+ 2H* + 2eEquation 1: Oxidation Reaction Once fuel activation has occurred, the total number of electrons and activation site surface area determine the maximum available current. The surface area isfixed by the physical design of the electrolyte electrode interface, which can be designed and therefore customized for agiven application. The total number of electrons however isdriven by the concentration of fuel accessing the interface. This concentration isdetermined by the type of fuel, and flow rate across the boundary. Although these characteristics are also constrained by design, to achieve a stable power level, fuel concentration must be maintained by either an automatic flow control system or meticulous operator control. Fuel that is not activated iscascaded away from the anode and may either activate at a farther downstream triple phase boundary location, or be pushed eventually to the exit of the fuel cell. This makes adequate flow control extremely important to maintaining a high level of efficiency. After the electrons and charge carriers have reached the fuel cell outlet location they are recombined at another triple phase boundary using a cathode and reactant. With oxygen as the reactant, recombination creates water according to the basic reduction reaction shown in Equation 2. 02 + 4H* + 4e- -> 2H 20 Equation 2: Reduction Reaction Because the overall reaction isathermodynamically favorable electron process and not achemical reaction, interactions occur without a release of free charge but produce heat, pure water, and power in 15 the form of electric current, as shown in Figure 2. This dictates that the reaction rate efficiency is limited by the amount of energy required to cause activation, recombination, and overcome several other inherent kinetic losses which are discussed in chapter 2.2. Ekmmmmmy e e Excess fuel H+ H2 H20 I+ H+ 02 Fuelin Airin Anode Elecrolyte Cthode Figure 2: Fuel Cell Flow Paths (O'Hare 2009) 2.2 Inherent Fuel Cell Losses Using the oxidation and reduction equations shown in chapter 2.1, the available electric potential for a fuel cell utilizing pure hydrogen isbounded by the idealized Gibbs free energy of 1.23 volts. When fuel cell kinetic limitations are levied on the system however, irreversible losses lower this value. The three major categories of losses which are all described here are: 1) Activation losses driven by the electrochemical reaction and cell composition 2) Ohmic losses affecting both ionic and electron conduction 3) Concentration losses due to mass transport inefficiencies The first type of irreversible fuel cell losses are activation losses. These losses are most significant at low current densities, but tend to stabilize once a large enough current density has been developed as shown in Figure 3. Free Energy 1.4 -Gibbs 1.2 - Activation Losses 1.0 4) - 0.8 - 4) 0.6 0.40.2- 0.5 1.0 1.5 2.0 2.5 Current Density (A/cm 2 ) Figure 3: Fuel Cell Activation Losses Activation losses affect low current operations because of how the fuel cell operates when it first comes into contact with fuel. When initially supplied with hydrogen at the anode and oxygen at the cathode, the system attempts to naturally balance the chemical reaction rate. Both oxidation and reduction reactions are free to occur simultaneously in the forward and reverse directions as triple phase boundary points are exposed to reactants. Initially the forward reaction rate at both electrodes islarger due to the electrical and chemical differential. Once the electric and ionic charge has built up on the electrode and within the electrolyte, the system has reached electrochemical equilibrium. This level is the Galvanic charge. With the system in equilibrium, further forward reactions are now no longer possible meaning no useful current isyet available. To forward bias the fuel cell and derive electric power, a decrease in cell voltage must be induced. This in effect drives the chemical reaction to the forward direction, but becomes adirect loss on output power. These losses can be mitigated though the selection of fuel cell material including to a large extent the type of catalyst selected. This occurs because a catalyst, platinum for example, makes the electrons more likely to break away from the fuel when it comes into contact with the electrode, so the electrochemical equilibrium occurs more easily. This effect can be recognized as a decrease in the activation energy plateau, meaning asmaller decrease in electric potential must be induced to forward bias the cell. During operation maintaining adequate activation energy plays a key role in developing the voltage current relationship. Ohmic losses are the second type of irreversible fuel cell losses, which are a linear function of current. Because electron transport isdriven by avoltage gradient across the electrode, and hydrogen ionic transport isdriven by both voltage gradient and chemical diffusivity through the electrolyte, any resistance to these flows will take away from the efficiency of the fuel cell. The losses exhibit a linear shape across the useable current ranges as shown in Figure 4. Gibbs Free Energy 1.4- Ohmic Losses > 0.4 02< 0.5 1.0 1.5 Current Density (A/cm 2.0 2 2.5 ) Figure 4: Fuel cell Ohmic Losses Fuel selection and operational conditions such as temperature, atmospheric pressure, and humidity affect the magnitude of ohmic losses. They are most dependant however on the amount of fuel cell 18 loading. Designers ultimately manage ohmic losses in the operational current regions through design considerations to reduce the electric resistance of the fuel cell electrodes while the ionic conductivity properties of the electrolyte are increased. The final type of irreversible fuel cell losses considered here are concentration losses. The basis for these losses are fuel transport inefficiencies, which leave the fuel cell either underpowered by not supplying enough fuel, or lead to inefficient use as too much fuel becomes exposed to the entire region of triple phase boundaries leaving fuel unused. To combat these losses designers can choose to install a constant stoichiometry flow control. Additionally, pure oxygen can be selected as the reactant vice air. This characteristic implies that for agiven fuel cell design, efficient use in the respective high current region isseverely limited. The impact concentration losses have on fuel cell performance isshown in Figure 5. 1.4- Gibbs Free Energy 1.2 1.0 - Concentration Losses 0.80.6 0.40.2- 0.5 1.0 1.5 2.0 2.5 2 Current Density (A/cm ) Figure 5: Fuel cell concentration Losses Once all the categories of losses are considered, the fuel cell's voltage to current profile isdeveloped by lowering the Gibbs free energy at each respective current as shown in Figure 6. 1.2 4) 0.8 0.6 0.4 0.2 0.5 1.0 1.5 2.0 2. Current Density (A/cm2) Activation Losses dominate Ohmic losses dominate Concentration losses dominate Figure 6: Fuel Cell Voltage Current Relationship with Irreversible Losses This nonlinear relationship ischaracterized by the Butler-Volmer relationship and can be used to determine fuel cell performance under a wide range of loading and atmospheric conditions. One advantage from a design perspective to quantifying how the electrochemical reaction rate affects ultimate fuel cell operation isthat output performance becomes an extremely controllable process. Because the factors which limit this rate have been meticulously characterized in industry, the efficiency of multiple types of fuel cells isalso known. This means that simple design modifications can be readily implemented where feasible to improve kinetic performance including: 1) Increasing reactant concentration or using pure oxygen 2) Decreasing activation barrier by catalyst selection based on the operating environment. For example, selection of aplatinum catalyst for use within a low temperature fuel cell, or nonprecious metal conductors for use at higher temperatures. 3) Increase fuel cell operating temperature which increases ion conductivity and fuel mass transport 4) 2.3 Increase the number of reaction sites by creating roughness in the interface region Types of Fuel Cells Fuel cells are distinguished by the chemical reactions which occur at both the oxidation and reduction triple phase boundaries. By selecting an applicable fuel source and electrolyte material, the fuel cell components and operational conditions required to achieve optimal performance generally become fixed. This implies that although multiple combinations of fuel cell components capable of producing usable current exist, only afew configurations are preferable and warrant discussion. The three considered for this research are the solid oxide fuel cell (SOFC), the polymer electrolyte membrane fuel cell (PEMFC), and the phosphoric acid fuel cell (PAFC). 2.3.1 Solid Oxide Fuel Cells (SOFC) The first type of fuel cell considered isthe SOFC. This fuel cell utilizes asolid oxide electrode which is typically constructed of yttria-stabalized zirconia and operates at temperatures of 600*F to 10000 F. Although these fuel cells possess adesirable operating efficiency, their relatively high operating temperatures require significant heat removal or cogeneration support systems, and drive up the cost as temperature resistant components are required. The major advantage of the SOFC operating temperature range isthat inherent losses are reduced and the fuel cell functions at a high efficiency. The temperature range also allows for selection of a nonprecious metal catalyst. Adding this type of selection flexibility mitigates some of the cost associated with a more robust design and manufacturing requirement, as well as adds significant poisoning resistance. Another advantage of the SOFC isthat it operates with a high level of fuel selection flexibility. If onboard reformation of diesel generator or gas turbine fuel isnot the most practical method of producing an acceptable reactant, the SOFC can readily be supplied with another more accessible fuel with only minor adjustments to the overall component architecture and support systems. The main problem for implementing SOFC systems onboard surface ships isthat materials which come in contact with high temperatures must be precisely designed and manufactured. This concern makes fabrication expensive and creates thermal cycling difficulties. To ensure each component's structural integrity isin fact maintained, the start up and operational cycling rates must be strictly regulated. Additionally, selecting the thickness of the electrolyte becomes problematic. To ensure ionic transport isnot impeded, the electrolyte must be as thin as possible. To maintain structurally integrity for meeting durability concerns however, the thickness must be increased. This contradiction means a SOFC may have either performance or durability issues for use within a given application. 2.3.2 Polymer Electrolyte Membrane Fuel Cells (PEMFC) The PEMFC operates by passing hydrogen that has been stripped of its electrons through a perfluorinated polymer. Electrons which have been removed from the fuel then bypass the electrolyte according to the process and reactions described in chapter 2.2. This configuration permits operation at extremely low temperatures compared to other fuel cells. Some advantages of PEMFC operating at such low temperatures isthat the fuel cell doesn't require as meticulous of a support system infrastructure which allows the fuel cell to be packaged in a compact manner. This makes the PEMFC extremely scaleable when space and volume limitations are not strict. In addition to these characteristics, the key benefit to PEMFCs from a naval architecture perspective, as previously described, are their lack of moving parts, and ability to produce only useable electricity and pure water. Because of the fuel cells low operating temperature requirement and the proton conducting ability of the polymer membrane, electrolyte thickness can be as low as 20 pm, which significantly decreases ionic conductivity losses. "The PEMFC currently exhibits the highest power density of all the fuel cell types (300-1000 mW/cm 2)." (O'Hare, 2009 p.264) Additionally, PEMFCs are able to quickly start and cycle across useable current levels, making it operationally flexible. One final advantage of using PEMFCs over other types isthat current research employed by auto manufacturers isfocused on component level upgrades that will tend to raise overall system level reliability, as well as increase operating efficiency even further. Although the PEMFC possesses attractive operating properties, there are several disadvantages to overcome when considering them as a valid power source. The first isthat the low temperature requires a platinum catalyst to mitigate activation losses, which drives up construction costs and create specific durability concerns. The system is highly susceptible to carbon monoxide and sulfur poisoning. After operational periods where fuel cell components are exposed to these fuel byproducts, performance decreases, and the potential for catastrophic failure of either the electrolyte or some other component becomes real. The reason this disadvantage iseven more applicable within naval applications isbecause onboard reforming of either diesel or gas turbine fuel must be used, which both have high concentrations of sulfur. Another disadvantage isthat by selecting perfluorinated polymer as the membrane material, an active water management system must be designed and implemented in order to maintain the electrolyte's integrity. If the membrane becomes damaged due to an improper water level, the fuel cell will in effect short circuit. This occurs because hydrogen molecules at the anode are allowed to directly interact with the cathode, which transfers electrons around the electric load instead of through it. Both of these disadvantages must be taken into consideration if an appropriate surface ship propulsion system isto be designed using PEMFCs. 2.3.3 Phosphoric Acid Fuel Cells (PAFC) The final type of fuel cell considered uses a phosphoric acid electrolyte in liquid form. This is accomplished by maintaining a reservoir of H2PO4 electrolyte material between two graphite electrodes. Due to the chemical characteristics of the electrolyte, the operating temperature of the fuel cell must be maintained between 100*F and 4000 F.This fuel cell configuration, similar to PEMFCs, transfers protons across the electrolyte which means it follows the anode and cathode reactions and limitations described in chapter 2.1. The primary advantages of this technology are that it ismature and currently employed in land based utilities as both primary and backup power sources. Because the operating temperature isslightly higher than a PEMFC, the efficiency isslightly higher but without all of the temperature dependant disadvantages seen in with the SOFC configuration. Asignificant detractor toward putting PAFCs onboard surface ships isthat the electrolyte isextremely corrosive and must be constantly monitored and replenished during operation. This requires that ship designers put appropriate storage and handling considerations into place. 2.4 Fuel Cell Selection The advantages and disadvantages of the three fuel cell types are summarized in Table 1. SOFC Disadvantages Advantages Fuel Cell Type -Fuel Selection Flexibility -High temperature material issues -Nonprecious metal catalyst -Rigorous support system requirements -High power density PEMFC PAFC -Good start stop capability -Platinum catalyst -Low temperature operation -Scalable power levels -Polymer membrane expensive and has potential for damage -Susceptible to Co and S poisoning -Mature technology -Corrosive liquid electrolyte -High reliability -Platinum catalyst -Low cost electrolyte -Susceptible to Co and Spoisoning Table 1: Fuel Cell Characteristics Comparison (O'Hare 2009) Although they are awell researched technology, due to the low flexibility and rigorous support system requirements, the SOFC isnot the most favorable selection for a naval application. When considering a PAFC for use within a naval architecture application, the additionally required corrosive material measures to support the phosphoric acid electrolyte, as well as the disadvantages already seen by PEMFCs of susceptibility to carbon monoxide and sulfur poisoning make this type generally less desirable than PEMFCs. 2.5 Durability Concerns Durability isone additional hurdle to be overcome before the PEMFC can be confidently used in a naval engineering application. To provide reliable power for the ships service electric plant and propulsion, the fuel cell and all of its supporting equipment must run for a large number of hours at the expected performance level. The multiple failure mechanisms leading to a power drop off, as well as any gradual performance degradations due to individual component malfunctions or improper operation must be 25 fully understood and ultimately controlled before fuel cells will serve as arealistic means of shipboard power generation. As afinal design consideration, the fuel cell must be designed in such a manner to compensate for any expected life cycle degradations. The first area that failure can occur in a fuel cell isacross the electrolyte. Because of the possibility for extremely thin membranes in the PEMFC it becomes an even more real concern. The electrolyte is susceptible to harmful mechanical interactions consisting of foreign material or substances from preexisting flaws which become worse and damage the physical membrane. Because electrolytes are designed to attract ions so readily, chemical and ionic attacks are another likely cause of degradation. This isdue to chemicals in the form of caustic free radicals and other harmful ionic contaminants become absorbed and performing unpredictable wear and damage. The second area that ishighly susceptible to durability issues within the PEMFC isthe electrode catalyst boundary. Afuel cell's output power is afunction of electrode surface area, which determines the rate at which electrons can be in effect absorbed from the fuel into useable current. Multiple chemical processes such as platinum pitting, oxidation, and dissolution, tend to decrease that surface area and will over time lower total cell efficiency. This implies that after platinum degradation has occurred, a higher amount of fuel will be required at the respective triple phase boundary to sustain the electrochemical rate that the same fuel cell experienced earlier in its life. There are still many challenges toward understanding each failure mechanism and how it relates to overall fuel cell kinetics. The advantage for naval engineering however isthat knowledge can be leveraged from the automotive industry, which has set future performance and reliability goals for the PEMFC. Automotive applications are similar in that although the power required to supplement shipboard propulsion isseveral magnitudes of order larger, the operational profile issimilar between the two applications. Car power sources must be able to start quickly from acold condition as well as operate in either a stop and go fashion or produce a sustained power level for a given period, and perform at power for a comparable life cycle. Until PEMFC reliability thresholds have been proven in less than ideal atmospheric conditions, even though fuel cell possess desirable powering and efficiency characteristics, they will not be acceptable for shipboard use. 3. Propulsion Plant Model Now that fuel cells as an alternate power source and PEM fuel cells in specific were selected as a desirable alternative power source for use within a hybrid surface ship propulsion plant, a propulsion plant optimization model was developed. The goal of the model was to determine the amount of fuel cell power that should be installed on a surface ship application to derive the largest decrease in fuel consumption. The program also determined how best to operate those fuel cells in concert with gas turbines. Describing the optimization model isconducted in two steps. The first isthe set of specific assumptions which were made with respect to the desired propulsion plant and ship operational characteristics. The second isthe general flow based algorithm, and resulting MATLAB* code. 3.1 General Assumptions Several general assumptions were made to assist in developing the code algorithm. These assumptions fed the model by constraining inputs and served to conservatively bind the optimization program. 3.1.1 Gas Turbine Inputs The first assumption was that a DDG-51 flight 1 US Navy destroyer isthe ship under analysis. This decision led to the first input, which isthat the propulsion plant for comparison isoutfitted with 4 LM2500 gas turbines dedicated to developing speeds of 35 knots through a mechanical drive propulsion system. The next input required by the program isthe useable gas turbine power range. Because all 4 turbines are capable of producing power in concert and each iscapable of developing 25 MW of comparable electrical power, the range of powers to be analyzed isup to 100 MW. This will serve as the maximum value for iteration when propulsion plant loading isconsidered by the model. Another assumption isthat the specific fuel consumption extends continuously across all possible loading states of the prime mover but is nonlinear. Afuel consumption curve used by this model was developed by scaling ageneral gas turbine consumption curve in to the 25 MW limit, shown in Figure 7. Power RAMW) Figure 7: Gas Turbine Specific Fuel consumption (Woud, 2002) ............. ---- .-J ..... Based on the shape of the specific fuel consumption curve and some additional analysis, the next assumption pertains to how the gas turbines are loaded. It isseen from the two total fuel consumption curves in Figure 8 that for a majority of the desired power ranges, the optimal fuel efficiency at a given load isachieved when gas turbines are operated sequentially. This occurs because bringing multiple gas turbines online in a parallel fashion keeps each one operating at a much lower power level forcing the specific fuel consumption of each higher. 0.9- 0.8 - Parallel Operation 0.7- 0.5- . E 0.4 -0.3- Series Operation 0.20.1 0 10 20 30 40M 0 P-we Axin (MW) 70 80 100 Figure 8: Total Fuel consumption of 4 Gas Turbines This assumption affects the optimization model as any benefit from adding fuel cells to the propulsion plant must be shown when compared to the best possible fuel consumption, which occurs under 4 gas turbine series operation. For loading purposes, all required power below 25 MW will be assumed by one prime mover. Another gas turbine will then be brought online to accept additional power requirements, while the original gas turbine continues operating at full load. Although this isnot the traditional operating pattern based on the current US Navy fleet tendency of having all gas turbines lightly loaded to provide high speeds in ashort notice, this process produces the best possible fuel use profile which will be benchmarked. Evaluating the fuel consumption differential curve in Figure 9, the gas turbine operating method assumption can be seen as valid for all values with the exception of powers between 75 - 85 MW. The overlap in this segment isdue to the starting cycle of the final gas turbine as it picks up load just over the 75 MW value. This forces it to operate at a very high specific fuel consumption, while the 4 gas turbines in parallel are by this point operating at more than 75% of their rated load. It isassumed that the magnitude of the fuel consumption represented by the overlapping section issmall and series operation will be considered here. 0.3- 0.25 - 02 - 0.15- 0.1 - 0.05 - 0 0 10 20 35 4 0 1 0 90 90 100 Figure 9: Fuel Consumption Differential Based on Operating Method Acorollary assumption concerning gas turbine operation was that even though they operate more efficiently under higher loading, it isnot beneficial to overload a gas turbine in hopes of achieving lower fuel consumption by decreasing the prime mover's specific fuel consumption. Because the shape of the gas turbine total fuel consumption curve does not experience a negative slope across any two power ranges, it ismost conservative for fuel usage to exactly load the prime mover at a desired level. This is seen discretely in Figure 10. If only 5 MW of power isdesired, a single gas turbine should not be operated at 10 or 15 MW. ................. Fuel Consumed at 5 MW = 5 MW x 475 grams per kwh = 2,375 kg/h Fuel Consumed at 10 WN 10 MW x 300 grams per kwh 3,000 kg/h A Fuel Consumed at 15 MW 15 MW x 250 grams per kwh= 3,750 kg/h 300- 2WC 0 1520 510 25 Powu Range(MW) Figure 10: Gas Turbine Fuel Consumption at 3 Loading Conditions 3.1.2 PEMFC Inputs Once assumptions and inputs for gas turbine operation were determined, identifying how the PEMFC operates in the hybrid propulsion plant was accomplished. The first assumption in this area isthat blocks of 250 kilowatt PEMFCs will be installed and operated with an automatic fuel control system. The propulsion plant will iterate from 1 to 40 of the fuel cells which allows up to 10% of the total required propulsion load to be powered either partially or fully from the alternative power source. To ensure even wear on fuel cell components, all installed fuel cell blocks are to be loaded to their desired power level simultaneously. This assumption isvalid from a fuel consumption perspective as efficiency islinear when a family of fuel cell voltage to current curves isanalyzed, shown in Figure 11. 0 2 4 6 8 10 Current Density (A/cm 2) Figure 11: Fuel Cell Voltage Vs. Current Density Using values in the plot provided, the power density can be seen as a linear function of the fuel rate shown in Table 2. Fuel Rate Cell Voltage (%flow) Volts Current Density A/cm2 Power Density Watts/cm2 Power Density / Fuel Rate 20 0.5 1.9 0.95 4.75 40 05 3.8 1.9 4.75 60 0.5 5.7 2.85 4.75 80 0.5 7.6 3.8 4.75 100 0.5 9.5 4.75 4.75 Table 2: Fuel Cell Operating Parameters For controlling the fuel cell, it was assumed that an accurate fuel control system was available and could accurately control power output in 2.5 kilowatt increments, or 1%of total fuel cell output power. To adjust for the finite timing and accuracy limitations that exist in current PEM fuel cells, 10% degradation to the assumed efficiency was inserted to add a level of conservatism. Fuel consumptions of 250 grams for each kilowatt hour could be seen in a real world application. With the efficiency decrease ................. implemented however, selection of PEM fuel cells in their current state yield a power source operating at an efficiency of 275 grams of fuel for each kilowatt hour. 3.1.3 DDG-51 Flight 1 Inputs The final set of assumptions has to do with the naval architecture application. The DDG-51 Flight 1 hull requires acertain thrust to achieve agiven speed. These values are known and are given for usable speeds in Figure 12. 120 100 80 5 10 20 25 15 SHIP SPEED, KT 30 35 Figure 12: DDG 51 Flight 1 Powering Curve The final assumption isthat adiscrete speed profile will be used. This assumption isvalid because in typical operating conditions, engine orders are applied in standard bell formats. For example an allahead 1/3 or full bell correspond to exactly 5 or 20 knots. The profile to be used by the optimization model isgenerated in the ASSET* 5.3 surface combatant mission speed probability array and is identified in Table 3. Speed (Fraction) Speed (Knots) Probability 5 0.119 0.167 * 0.500 * Vsustained 15 0.466 0.667 * Vsustained 20 0.356 0.833 * Vsustained 25 0.044 1.000 * Vsustained 30 0.015 Vsustained Table 3: DDG-51 Flight 1 Mission Speed Profile 3.2 Model Algorithm and code Although developing the algorithm was an iterative process, the final version was completed after all of the general assumptions and inputs were made. This version, shown in Figure 13, identifies the desired order and progression of the program. Inputs Process Nunter of gasturtinesinstad Useable gasturbine power range Gas turbine SFC Bst gin turbine operatin schierne Typeof fuelcel nudeled useabe fuece owrrag Fuel cel SFC sPs resistance cwve SNpsspeedpronie Figure 13: Propulsion Plant Optimization Algorithm 34 Outputs Determining the minimum fuel consumption during a DDG-51 underway as afunction of number of installed fuel cells was the ultimate goal of the optimization program. Achieving this was accomplished by first establishing and iterating the states of three program variables, identified on the top left in the algorithm process section. The first program variable, and where the model begins, isused to increment the amount of fuel cell capacity installed within the hybrid propulsion plant configuration. This variable has the most significant ramifications on the overall naval architecture of the ship design. This occurs because the required support and control systems for the fuel cell, as well as the physical block of fuel cells all take up propulsion plant surface area. They also add a large amount of weight to the ship's propulsion spaces. Although it has not yet been shown as a feasible quantity from an installation and operation perspective, it was assumed that up to 10% of the ships propulsion requirement could be assumed by fuel cells. This assumption requires the fuel cell capacity variable to iterate 40 times, increasing by one until all 40 configurations have been evaluated. Taking the range into consideration, and based on the optimization model's results, avalue for the installed capacity will ultimately be selected and atotal ship design will be developed and briefly analyzed for technical feasibility. The second variable introduced and required for optimizing fuel consumption was propulsion power. This value was iterated from 0 to 100 MW in 100 kilowatt increments for each of the possible propulsion plant configurations, with 100 MW being the power required to propel the ship to near 35 knots. Although this power level iscurrently achieved by the four LM-2500 gas turbines, if any power can be more efficiently met at a lower fuel consumption rate with partial load taken by the fuel cells it needs to be considered by the model. The third variable was used to set the operating status of the installed fuel cell capacity. This is necessary because even though a fixed fuel cell configuration ispreviously established and placed under load, the model should not be constrained into having to operate the fuel cells at 100% of their rated capacity for a given propulsion plant loading requirement. Iterating this quantity was done in 1%fuel cell increments, or 25 kilowatts. The operating status will ultimately be selected which performs at an overall lower fuel consumption compared to others, including the gas turbines only value. Although multiple operating conditions could exist for agiven propulsion plant configuration and load that each consume the same amount of fuel, the one with the lowest number of operating fuel cells will be chosen. The next function of the optimization model after the state of each of the three variables has been established isto determine the overall fuel consumption of the hybrid propulsion plant. This value was determined in a three part summation equation. The first part sums the fuel consumed by all fully loaded gas turbines, which isthe product of the number of fully loaded turbine sets, gas turbine power range, and the specific fuel consumption of the gas turbine at its fully loaded state. The second part is used to determine and sum the fuel consumed by the installed fuel cells. This isfound by determining the product of operational fuel cell capacity and fuel cell specific fuel consumption at the respective power level. The final segment in the fuel consumption equation isthat which isassumed by a partially loaded gas turbine. To get the amount of power that is in a partially loaded state, the power assumed by fully loaded gas turbines and operating fuel cells issubtracted from the total power required. This partially loaded quantity isthen multiplied by the specific fuel consumption of the gas turbine while operating at the appropriate ratio of its full load. Once fuel consumption isfound it iscompared to the gas turbine only as well as the current optimal value, and a new optimal value ispotentially set. The routine then continues iterating until it has evaluated all program variable combinations. After all states have been examined, a family of optimal fuel consumption curves as a function of required power isoutput for each of the installed hybrid plant configurations. These consumption curves are then converted based on the ships speed profile and resistance curve into an overall fuel consumption which is now only afunction of installed fuel cell capacity. 4. Optimization Results Once the optimization model was translated into MATLAB* code, the outputs defined in chapter 3.2 were generated. This chapter explains those results, as well as highlights the potential reasons for any trends or unexpected outcomes. There isalso a sensitivity analysis conducted on the model parameters for speed profile and fuel cell specific fuel consumption. 4.1 Model Output The first output from the model isthe family of total fuel consumption curves, which are functions of required propulsion power. Each curve in the figures below represents a different propulsion plant configuration, from 1to 40 fuel cells installed. The plot in Figure 14 shows required power up to 10 MW. 0.16- Gas Turbine Only Configuration 0.14- 15 knots 0.12- 0.16E l3 5 knots u. 0.04 - 20 Fuel Cells 0~~~~3 I I 7 8 0.02 - Fue Cellsel' 10 Fuel Cells 01 2 3 4 5 6 Powr Range (MW) 9 10 Figure 14: Total Fuel Consumption up to 10 MW The figure also displays the gas turbine only curve, which represents the baseline propulsion plant with no fuel cell capacity installed. This gas turbine only curve serves as a de facto ceiling for optimized fuel consumption as any regions of required power where the best fuel efficiency isavailable with only gas turbines should be fulfilled with this operating method. In these scenarios, regardless of the installed fuel cell capacity, the operating status of all fuel cells will be to the off position. Another plot of total fuel consumption for up to 100 MW isdisplayed in Figure 15. The discrete speeds considered by the ship speed profile are noted on this total fuel consumption plot as well. ......... ... .......... ................................. 0.9 0.8 - 0.5 - 25 knots 0.4 - 20 knots 02-15 Inefficiency region caused by gas knots turbine startup cycle 0.1 0 10 20 30 40 so Power Range (MW) s0 70 so g0 100 Figure 15: Total Fuel Consumption up to 100 MW This curve shows that there are 4 finite regions within the power range domain where the gas turbines operating alone produce a relative inefficiency. Using both the 10 MW and 100 MW plots, a total of four out of the seven nominal engine order speeds are seen to fall within these regions. This implies that installing fuel cell capacity will reduce overall fuel consumption to the extent that these speeds are favored over speeds not within the inefficiency regions. The next output from the optimization model isthe difference in total fuel consumption between the gas turbine only curve and each of the 40 hybrid plant curves. The difference curves show that the regions along the power domain where fuel cells provide a consumption benefit are in fact discretely located. Each difference peak increases rapidly and drops after several megawatts of load have been placed on the partially loaded gas turbine, shown for asingle pulse in Figure 16. 20 Fuel Cells 10 Fuel Cells Cells 5 Fuel Cells - 40 Fuel Cells I Fuel Cell Figure 16: Fuel Consumption Differential for a Single Inefficiency Region This fuel consumption difference curve isalso shown in Figure 17 across the entire 100 MW region as a family of curves for each of the 40 propulsion plant configurations. The four pulses are seen to start coincident with each of the four gas turbine startup cycles as expected. They then taper off in favor of the gas turbine only profile when about 10 MW of load isable to be placed onto a partially loaded gas turbine. This process isseen to correlate with the gas turbine specific fuel consumption reaching the much lower slope in Figure 7 of Chapter 3.1.1. 0.040.0350.03 0.025 0.02 a .2 E C0.01 u- 0 .0 0 5 *6 0 10 20 30 70 Power Ranoe (MW) Figure 17: Fuel Consumption Difference Family of Curves 80 90 100 The final output from the optimization model is produced when the general fuel consumption curves as afunction of power required are converted into total fuel consumed for a definite ship operating timeframe. This isaccomplished within the optimization model by taking the family of total fuel consumption curves and considering the power required for each speed as well as the probability that the ship operates at that speed. The result isdisplayed in Figure 18 as a function of the number of installed fuel cells. Gas Turbine Only Total Fuel Consumption 0.98- 8 0.96- 0.94- 9 S0.92- Total Fuel Consumption as Function of the Number of Installed Fuel Cells -A 0.9- 0.88 0.86 - 0.84 0 5 to 15 20 25 Number of Installed Fuel Cells 30 35 40 Figure 18: Total Fuel Consumption for a DDG-51 Speed Profile These results show that there islittle decrease in fuel consumption, on the order of 3%, until around 12 fuel cells or 3 MW of capacity, have been installed. At this point, total fuel consumption decreases rapidly for each additional fuel cell installed until a 20 bank fuel cell, totaling 5 MW of capacity, has been reached. At this point, the propulsion plant has experienced a 14% decrease in fuel consumption. Increasing the number of fuel cells past this point however does not continue lowering fuel consumption. An ancillary program output isthe operating status of the fuel cell bank required to achieve the optimal fuel consumption at each required power. The fuel cell operating status for the 4 propulsion plant configurations with 5, 10, 15, and 20 fuel cells for up to 10 MW isdisplayed in Figure 19. 0.8 0.7 Z.6 3 0.5 0. 4 ~0.3 LL 10 cells 5 cells o.1 15 cells 0.1 20 cells 0 10 Figure 19: Fuel cell Operating scheme up to 10 MW The gas turbine startup cycles are again seen to drive the desired fuel cell operation as 4 similar operating patterns are observed at each of the regions where gas turbines hold a low partial load, seen for a 100 MW range in Figure 20. Operation of the fuel cells from configuration to configuration isseen to be similarly timed. All four plant configurations are seen to pick up the entire initial propulsion load and operate fully loaded until about 5 MW of additional power isrequired. At this point each plant begins shifting over to a partially loaded gas turbine. When the 10 MW partial load value has been reached, the gas turbine isnow operating efficiently enough to accept the entire load and all 4 fuel cell banks are turned off. 0.9 0.8 0.7 0.6 0.5 0.3 0.2 0.1 0 0 20 30 40 50 60 Pw 2 Fmed (MW) 70 0 90 100 Figure 20: Fuel Cell Operating Scheme up to 100 MW Finally, Table 4 isproduced which shows for the 20 fuel cell configuration how best to operate each piece of propulsion plant machinery. An important item in this table isthat for the first 3 speeds, the power required can be generated almost solely by the fuel cell bank. Although this requires meticulous fuel and reactant gas flow control, because the power levels are discrete, an automatic flow control system for quickly reaching the required output level can be used. The system would then require slight feedback control for fine tuning the power required during a several minute transient response. Another item of note isthat for all speeds less than 25 knots, the fuel cells and a single gas turbine are sufficient for meeting any power demand. 4.2 Sensitivity Analysis Now that the optimization model has been coded and simulations run, it is important to investigate some of the initial assumptions. The goal of this chapter isto increase the confidence in the models outcome by altering 2 variables and determining their significance on the final results. Sensitivity analysis in this manner also develops a focus for making recommendations on future areas of study. 4.2.1 Varying the Speed Profile The first area looked at more in depth iswith respect to the speed profile. Because four speed profile values fall within a gas turbine startup cycle region it was readily expected that fuel savings could be attained. Knowing how significant an impact the given speed profile on the savings in total fuel consumption will be found by shifting the probabilities from four speeds which fall within gas turbine inefficiency regions to those which can already be achieved in an efficient manner. This will be accomplished by decrementing the probability of operating at inefficient speeds while increasing the time spent at more fuel efficient speeds. Figure 21 represents the speed profiles used to conduct this analysis. 0.50 0.45 m kiitial Speed Profle 0.40 M5%Shift To Bficient Speeds 0 = 0.35 10 a 10% ShIt To Efficient Speeds 0.30 015% Shift To Efficient Speeds 4 0.25 * 20% Shift To Bfficient Speeds 0.20 o UFinal Speed Profile 0.15 0.10 0.05 0.00 ... - 5 knots 10 knots 15 knots 20 knots 25 knots EAiar 30 knots 1 35 knots Ship Speed Figure 21: Decreasing Operation at All Inefficient Speeds The resulting total fuel consumption seen in Figure 22 is highly sensitive to the speed profile. By shifting the speed probabilities by only 5%, the entire gain of 14% fuel savings originally possible by the 20 fuel cell installation issignificantly impacted. As the speed profile isfurther shifted, the savings continue diminishing. 0.98- Shift by 25% to a high speed oriented speed profile Shift by 20%to a high speed oriented speed profile 0.96-Q 0 0.94- Shift by 15% to a high speed oriented speed profile 0.92- Shift by 10% to a high speed oriented speed profile 0.9 - 088- Shift by 5%toa high speed oriented speed profile 086 0.84 DDG-51 Speed Profile 0 5 10 1 15 25 20 Number of Installed Fuel Cells 30 35 40 Figure 22: Steaming Profile Sensitivity Analysis for all Low Speeds The next way to analyze the sensitive which speed has on total fuel consumption isto shift the speed profile away from asingle speed and toward the 3 speeds not affected by the gas turbine startup cycles. This was accomplished by performing 5 runs which removed 10% of the time spent steaming at 15 knots and replacing it by an equally distributed probability at higher speeds, shown in Figure 23. 0.50 0.45 | hIitial Speed Profile 0.40 10% Shift From15 Knots r 0 = 0.35 O 20% Shift From 15 Knots 0.35 0 0o S 0.25 30% Shift From 15 Knots 0 40% Shift From 15 Knots 0.20 MFinal Speed Profile cM 0.15 0.10 0.05 0.00 5 knots 10 knots 15 knots 20 knots 25 knots 30 knots 35 knots Ship Speed Figure 23: Decreasing Operation at a Single Inefficient Speed The results of this analysis, displayed in Figure 24, show that again the makeup of the speed profile has a very significant impact on the resulting total fuel consumption. The opportunity for fuel cells to be effective in developing savings therefore ishighly dependent on any changes the US Navy implements in how they operate this class of destroyer. 0.98- 0.96 - 0.94- ( i 0.92- Shift 15 knot probability by 30% to a high speed oriented speed profile E 0.9- Shift 15 knot probability by 20% to a high speed oriented speed profile ,0.88- to a high speed oriented speed profile 0.868 DDG-51 Speed Profile 0.84 0 5 10 15 20 25 Nmber of rtaied Ful Cells 30 35 Figure 24: Steaming Profile Sensitivity Analysis Shifting Away from the 15 Knot Speed 40 In summary, slight modifications to the speed profile can drastically affect the resultant total fuel consumption. Installing fuel cells isbeneficial to reducing fuel consumption but changes in the operational profile can cause drastic changes to this overall reduction, shown in Table 5. Sensitivity Analysis Modifying All Speeds Modifying 15 Knots Initial Speed Profile 14% savings 14% Savings 1st iteration 12% Savings 11% Savings nd Iteration 10% Savings 9%Savings 3rd Iteration 9% Savings 7% Savings 4th Iteration 8% Savings 5%Savings Final Speed Profile 8.5% savings 4% Savings 2 Table 5: Summary of speed Profile Sensitivity Analysis 4.2.2 Varying PEMFC efficiency The second area sensitivity analysis iswarranted iswith respect to the specific fuel consumption of the fuel cell. Although a275 gram per kilowatt-hour quantity was assumed, because research in this area is ongoing in concert with the automobile industry, advances may be made. This has to potential for making the fuel cell more efficient. Conversely, if avalid hydrogen infrastructure isnot enacted in the near future, and onboard reforming of traditional engine fuels proves more challenging than currently expected, this value may be much lower. Accordingly, total fuel consumption as afunction of installed fuel cell capacity was determined for fuel cell specific fuel consumptions in 5%increments as shown in Figure 25. .............. - - ................... . .............. ............................ 0.90- '& 0.94- 0.92- Fuel Cell SFC = 302.5 0. 8 (A Fuel Cell SFC = 288.8 0.88 *Fuel Cell SFC = 275. Fuel Cell SFC = 261.3 S0.86- Loss In Fuel 2%Gain in Fuel Savogs 2%Gain inFuel Savings Fuel Cell SFC = 247.52% . *SFC of 275.0 is the baseline model value 0.8 0 5 10 15 20 25 30 35 40 mbr ofWi staled FuelCels Figure 25: Consumption Sensitivity based on Fuel Cell SFC Compared to the results of the speed profile sensitivity analysis, total fuel consumption appears relatively stable when exposed to changes in the fuel cells specific fuel consumption. For each of the 5% SFC increments, increasing or decreasing, the fuel consumption curves exhibit an extremely tight shape and show little deviation with a low number of fuel cells installed. As the number of fuel cells increases, there isa2%deviation in total fuel savings for each increment, but the overall shape remains intact and istherefore predictable. 5. Hybrid Propulsion Plant Integration The results of the propulsion plant optimization program in chapter 4 have shown that installing 5 MW of PEM fuel cells have the potential for providing up to a 14% decrease in fuel consumption for the DDG51 Flight 1hull under a reasonable steaming profile. This chapter discusses the naval architecture ramifications of installing such a system shipboard by using the ship modeling tool ASSET* version 5.3 which iscapable of comparing a hybrid propulsion plant ship to one that isgas turbine only driven. The benefit of fuel cell use aboard US naval surface combatants isseen because they increase the overall fuel efficiency of propulsion equipment while operating at desired speeds. This benefit directly translates into less fuel used during a transit, or to an overall longer deployment as the ship now possesses increased endurance. Either of these options translates into less money being spent. These savings can also be manifest in a more indirect manner. If the hybrid fuel cell and gas turbine propulsion plant isimplemented at the early concept design phase, fuel tanks can be sized appropriately for the given mission, and a more optimized design can be developed. Early stage design of the hybrid propulsion plant would also allow for the fuel cells and supporting equipment to be arranged in an efficient manner. Characteristics such as survivability, maintainability, manufacturability, and operational considerations could also be designed for in this scenario. ASSET* is first used to analyze the gas turbine only ship, with the machinery arrangement of the DDG-51 shown in Figure 26. AP Il 1.0 0.9 I I I I I 0.8 0.7 0.6 0.5 0.4 0.3 I 0.2 0.1 I 0 20 FP I I | 40 60 0.0 SCALE M Figure 26: DDG-51 Machinery Layout In this study, fuel cell use can lead to 14% of the fuel tanks not being required to achieve the requisite 4000 nautical miles currently experienced by destroyers. DDG-51 designs dedicate 2,000 m3 of the ship's total 28,000 m3 in hull volume to fluid storage tanks. Of this amount, 1,522 M3, or just over 75%, is used for storing the ships 1,145 metric tons of fuel. By decreasing this amount by 160 metric tons, sufficient space becomes available to house the fuel cells and any additional hybrid plant machinery, including power generation, mechanical transmission, and chemical support components. The DDG-51 hull form primary characteristics are shown in Table 6. PRINCIPAL CHARACTERISTICS LBP HULL LOA BEAM, DWL DEPTH @ STA 10 DRAFT TO KEEL DVL DRAFT TO KEEL LUL FREEBOARD @ STA 3 GMT CP CX SPEED(KT): ENDURANCE: - M 142.0 153.4 18.0 12.7 6.2 6.7 7.4 0.9 0.587 0.825 MAX= 30.5 SUST= 29.3 3697.0 NM AT 20.0 KTS TRANSMISSION TYPE: MAIN ENG: 4 GT MECH @ 19220.4 KW SHAFT POWER/SHAFT: 37487.0 KW PROPULSORS: 2 - SEP GEN: WEIGHT SUMMARY - ETON GROUP 1 - HULL STRUCTURE 3275.5 GROUP 2 - PROP PLANT 761.3 GROUP 3 - ELECT PLANT 317.6 GROUP 4 - COME + SURVEIL 436.1 GROUP 5 - AUX SYSTEMS 856.7 GROUP 6 - OUTFIT + FURN 663.0 GROUP 7 - ARMAMENT 320.0 - CP 8 3 GT 24-HR LOAD MAX MARG ELECT LOAD 5.2 N SUN GROUPS 1-7 DESIGN MARGIN 6630.3 829.6 LIGHTSHIP WEIGHT LOADS 7459.9 1710.4 FULL LOAD DISPLACEMENT FULL LOAD KG: M 9170.3 7.8 MILITARY PAYLOAD WT- ETON USABLE FUEL WT - ETON DIA 1115.8 1145.5 2500.0 Kb 2377.1 3600.8 REQUIRED AREA SUMMARY - M2 MANNING ACCOM OFF 26 29 CPO 24 27 ENL 291 321 TOTAL 341 377 AVAILABLE AREA SUMMARY - M2 OTHER AREA - 5180. HULL AREA - 3995. SUPERSTRUCTURE AREA - 1161. SUPERSTRUCTURE AREA - 1896. TOTAL AREA - 6341. TOTAL AREA - 5891. REQUIRED VOLUME SUMMARY - M3 OTHER VOLUME 25072. SUPERSTRUCTURE VOLUME 2920. AVAILABLE VOLUME SUMMARY - M3 HULL VOLUME 22126. SUPERSTRUCTURE VOLUME 5403. TOTAL VOLUME TOTAL VOLUME - 27992. - 27529. Table 6: ASSET* 5.3 DDG-51 Design Summary To adapt the surface combatant hull, several modifications were made to the original DDG-51 model to allow for a hybrid propulsion plant. The first was that usable fuel weight was decreased by 14% to 985 metric tons. The design tool uses this parameter to develop all tanks, which it then places inside the hull. The locations of these tanks directly affect the placement of compartment subdivisions. By affecting the locations of transverse bulkheads, and other structural components, this step in the process plays a large role in determining the ship's large scale naval architecture parameters such as displacement, draft, and operating performance. Inthis case it isexpected that the lower required fluid volume will translate into slightly smaller ship, but because the fuel cells and supporting components are heavier, the ship will have a slightly larger draft and sit lower in the water. The next step to investigating a hybrid propulsion plant model for a surface combatant was modifying the machinery room spaces and layout. The ship's service generator located in the forward most auxiliary machinery room was shifted aft to provide deck space for the fuel cells. To emulate the effect that 5 MW of fuel cells and supporting equipment, including onboard fuel reformation, would have on the ship's architecture, a weight and space adjustment was made to the ships requirements table. Using the integrated propulsion system worksheet provided by the Electric Ship Program Office, 5 MW of fuel cells can reasonably be expected to require a 7 x 4.3 meter surface area profile and weigh 77 metric tons. Several support system items were also installed in this manner, which are summarized in Table 7. Length Width Height Removed from Ship Weight Area Volume 14% of Fuel Tank Capacity 160 MT 140 m 210 m3 Components Added Weight Area Volume Length Width Height 5.0 MW PEM Fuel Cell 77 MT 30.1 M 148 m3 7.0 m 4.3 m 4.9 m Onboard Reforming Plant 65 MT 18.5 M 46.2 m3 4.2 m 4.4 m 2.5 m Chemical Support System 8.4 MT 7.8 m2 13.2 m3 3.4m 2.3m 1.7m Mechanical Support Systems 54 MT 15.0 M2 5.0 m3 Group of Multiple Components Electrical Support Systems 32 MT 9.4 M2 3.1 m3 Group of Multiple Components Total Added 236.4 MT 80.8 m2 215.5 m' From Various Locations Table 7: Ship and Equipment Modifications Once the adjustments to the fuel tanks and machinery rooms were made within the model data base, the final step was to have ASSET* compile and synthesize the hybrid ship. To ensure ease of operation, the hybrid plant machinery layout is seen to be very similar to the DDG-51 flight 1 layout and isshown in Figure 27. Fuel Cell Power Unit Onboard Reformation Plant AP 1.0 FP 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 I I i I 0 10 20 30 0.1 0.0 SCALE M Figure 27: Hybrid Propulsion Plant Although the machinery layout doesn't reveal significant structural changes to the ship's architecture, there are notable changes to the overall geometry and principle characteristics, shown in Table 8. The ship has become significantly smaller, displacing 250 metric tons less than the DDG-51 hull. Both length and beam have also decreased, to 134.0 meters and 16.9 meters respectively. Also of note isthat the ships center of gravity has moved to lower in the ship by almost afull meter. This shows that fuel tanks were likely taken from relatively high in the ship while the more mass dense electric and mechanical components to be installed were placed low. Additionally, this new center of gravity has served to raise the ships meta-centric height from 0.9 meters to 1.2 meters, which means that the ship will exhibit a higher initial static stability. - M 134.0 144.7 16.9 12.7 6.2 7.1 7.0 1.2 0.587 0.825 PRINCIPAL CHARACTERISTICS LBP HULL LOA BEAM, DWL DEPTH 8 STA 10 DRAFT TO KEEL DWL DRAFT TO KEEL LWL FREEBOARD @ STA 3 GMT CP CX SPEED(KT): ENDURANCE: MAX- 30.0 SUST- 28.7 3094.1 NM AT 20.0 KTS TRANSMISSION TYPE: MAIN ENG: 4 GT MECH 8 19220.4 KW SHAFT POWER/SHAFT: PROPULSORS: 2 - CP SEP GEN: 37487.0 KW WEIGHT SUMMARY - HTON 1 - HULL STRUCTURE 3014.7 2 - PROP PLANT 1052.8 3 - ELECT PLANT 302.5 4 - COMM + SURVEIL 426.9 5 - AUX SYSTEMS 807.2 6 - OUTFIT + FURN 633.5 7 - ARMAMENT 319.9 SUM GROUPS 1-7 DESIGN MARGIN 6557.4 820.5 LIGHTSHIP WEIGHT LOADS 7377.9 1541.5 FULL LOAD DISPLACEMENT FULL LOAD KG: M 8919.4 6.9 MILITARY PAYLOAD WT- ITON USABLE FUEL VT - MTON 1115.8 985.0 5.2 M - 8 3 GT GROUP GROUP GROUP GROUP GROUP GROUP GROUP 2500.0 KW 24-HR LOAD MAX MARG ELECT LOAD 2265.6 3382.8 MANNING ACCOM OFF 26 29 CPO 24 27 ENL 291 321 TOTAL 341 377 REQUIRED AREA SUMMARY - M2 OTHER AREA 5376. SUPERSTRUCTURE AREA 1104. AVAILABLE AREA SUMMARY - M2 HULL AREA 3674. SUPERSTRUCTURE AREA 1599. TOTAL AREA TOTAL AREA 6480. - - 5273. REQUIRED VOLUME SUMMARY - M3 OTHER VOLUME 24755. SUPERSTRUCTURE VOLUME 2776. AVAILABLE VOLUME SUMMARY - M3 HULL VOLUME 19562. SUPERSTRUCTURE VOLUME 4566. TOTAL VOLUME TOTAL VOLUME - 27531. - 24128. Table 8: Hybrid Ship Principle Characteristics A side by side comparison of the ship's key characteristics is in Table 9. This table shows that even though fuel cells with all of their support systems considered are larger per megawatt of power delivered than gas turbines, because they create a large enough fuel savings, their installation can potentially serve to make the ship smaller and more stable. Hull DDG-51 Flight 1 Hybrid Surface Difference (%of Combatant Reference Value) Displacement 9,170 metric tonns 8,919 metric tonns 2.7% Reduction Length 144.7 meters 134.0 meters 7.4% Reduction Beam 18.0 meters 16.9 meters 6.1% Reduction Draft 6.7 meters 7.1 meters 6.0% Increase Meta-centric Height 0.9 meters 1.2 meters 33.3% Increase KG 7.8 meters 6.9 meters 11.5% Reduction Table 9: Principle Characteristics comparison Table 6. Summary Fuel cells are a nontraditional power source and create electricity in a clean and efficient manner. Although there are many types being researched and in operation in the commercial sector, PEMFCs present promise for use onboard US naval ships over the next 20 years. That being said, there are many technical hurdles yet to be overcome in the commercial realm before PEMFC's possess the robustness and power density to serve this function. Insummary, 5 MW of fuel cells when installed onboard the current DDG-51 hull form produce a 14% fuel savings. Although an optimization model has shown this fuel savings in Chapter 4, and aship modeling tool has displayed rough order technical feasibility in Chapter 5, there are substantial difficulties when attempting to install a nontraditional power source onboard an already designed destroyer. These include having to conduct detailed design within the machinery room and evaluate the ship's fuel tank layout to ensure the changes do not impede operation and maintainability. Additionally, the effects on ships performance will also have to be reexamined as there are changes to the ship's center of gravity. If installing 5MW of fuel cells onboard the DDG-51 hull form isnot possible, placing them in a gas turbine propulsion plant on another hull form isalso beneficial, but again presents challenges. Although the fuel savings results could be similar, the selected hull's resistance curve would have to be investigated to identify which speeds are in fact the most fuel inefficient and compare those results in 57 light of that hulls speed profile. Adifferent hull form with a new operational profile may mitigate some of the results shown in this research on the DDG-51 hull. Finally, if fuel cells cannot be installed onboard the current DDG-51 or any future gas turbine only surface ships, the operating status quo should be modified in an attempt to save fuel within the constraints of already existing machinery layouts. There are two ways this can be accomplished. The first isby operating gas turbines in series wherever possible. For a given engine order, this method will force more fuel efficient gas turbine operation. The second is by favoring speeds which keep all operating gas turbines as highly loaded as possible. The ship's resistance curve should be carefully analyzed and speeds avoided as much as possible which force a gas turbine to be loaded with less than 10 MW of power. 7. Bibliography O'Hayre, R.Cha, S.W, Colella, W., Prinz, F.B. (2009). Fuel Cell Fundamentals. John Wiley & Sons, New York. Spiegel, C.(2008). PEM Fuel Cell Modeling and Simulation Using MATLAB*. Academic Press, London. Woud, H. C., Stapersma, D.(2002). Design of Propulsion and Electric Power Generation Systems. IMarEST Publications, London. ........... ....... Appendix: Propulsion Plant Model Code The model developed from the algorithm described in chapter 3 is here listed as MATLAB* code. Results of the code in the form of output tables are found in chapter 4 of the main text. close all; clear all; hold on; gramstolbs=0.0022046224760379584; lbs_to_LT=0.00044642857; Electric-required = (1:10:100000); iInput Power Scurce Details: %Currently Installed Gas Turbines: GTavailable = 4; %Establish numbet of installed Gas Turbines GTsfc = (750 750 730 710 690 670 650 630 610 590 575 555 535 512 495 480 464 442 432 420 409 400 391 382 372 361 352 342 331 322 316.5 311 305.5 300 296 292 288 284 280 276 272 268 265 262 259 257 255 253 251 250 249 248 247 246 245 244 243.25 242.7 242.5 241.8 241.25 240.6 240 239.5 239 238.85 238.75 238.25 238 237.75 237.5 237.25 237 235.75 236.25 235.85 235.6 235.3 235 234.75 234.25 233.85]; GT-power range = 25000; %Establish relationship from kw power to GT Specific Fuel Consumption Array in g/kw*h GT installed = GT available * GTpower range; %GT kw installed/available plot (GTpower range /(1000*length(GT_sfc)) :GTpowerrange/ (1000*length(GT sfc)) :GTpower range/100 0,GT_sfc) xlabel('Power Range (MW)') ylabel('Specific Fuel Consumption (grams per kwh)') %Potentially Available Fuel Cells: FClinearsfc = 247.5; %CONSERVATIVE FC sfc EFFICIENCY. FCavailable = 4; %Number of fuel cells FCsfc = [FClinearsfc FClinear_sfc]; FC-powerrange = 1250; installed %Establish FC Specific Fuel Consumption Array %plot(O:FCpower range/length(FC sfc)+1:FCpoweriangeFC sfc) for n=1:FC available FCinstalled = n * FC power-range; %Assign FC kw installed/available for test Installedpower=GTinstalled+FCinstalled; %if GT_installed<Electric_required(length(Electric required)) ship %confirm GT alone can power % display ('not enough GT installed power'); %end %if Installed powercElectricrequired(length(Electricrequired) provides enough power % display ('not enough total installed power'); %end %confirm configuration FCactive=1; for l=1:length(Electric_required) %increment through each element of the electric required domain %Best use ratio(l)=O %BLOCK 1 Find most efficient operating conditions foi given installed power GTsrequired=fix(Electric required(l) /GTpower range); NominalGTload=Electric required(1) -GTsrequired*GTpowerrange; GTpartial-efficiency=fix(100*NominalGT-load/GTpowerrange); if GTpartialefficiency<=1 GT_partial-efficiency=1; end if GTpartialefficiency>=length(GT-sfc) GTpartial-efficiency=length(GT-sfc); end NominalGTefficiency=GT sfc(GTpartial-efficiency); %GT parallel efficiency two=GT sfc (+fix(GTpartial %if GT parallel efficiencytwo<=1 GT_parallel efficiency two-l; % %.end efficiency/2) ) GT_parallelefficiencyall=fix(length(GT sfc)*Electric required(l)/Electric required(length(Elect ric required))); if GT-parallel_efficiencyall<=1 GTparallel efficiencyall=1; end if GT-parallelefficiency all>=length(GTsfc) GTparallel efficiency all=length (GTsfc); end NominalGT_parallelefficiencyall=GT sfc (GT_parallelefficiencyall); Fuelconsumption (1,n) =GTsrequired*GTpowerrange*GT sfc (length (GT_sfc)) +NominalGTload*Nominal_ GTefficiency; GTonlyconsumption(l,n)=Fuelconsumption(l,n); %tabulate GT only fuel consumption at all powers - BASELINE %GT_parallel consumption two(l,n)= (GTs_required*GTpowerrange+NominalGT_load)*GT parallel effic iency_two; GTparallelconsumption all (1,n) =Electric required (1)*NominalGT parallelefficiencyall; %Find Combined fuel consumption and compare it to GT only value; for m=FC active:100; %increment loading of installed power to include fraction(m/100) of FC if m*FCinstalled/100<=Electric-required(l) %only increment FC loading if needed FC_loading(m)=m*FCinstalled/100; %Load n installed FCs to m fraction of loading GTloading(m)=Electric_required(l)-FCloading(m); %Power remaining to be supplied with GTs if GT-loading(m)<=0 %set zero GT load if GTloading(m)=O; end FC powering is sufficient %Assume GTs are brought online in series (ie MOST effeciently) %Find partially loaded GT Efficiency: GTsfullyloaded=f ix (GT loading (m) /GTpowerrange); PartialGTload (m)=GT loading (m)-GTs_fullyloaded*GTpowerrange; GToperation(m)=fix(100*PartialGTload(m)/GT_powerrange); if GToperation(m)<=1 GT_operation (m)=1; end if GT_operation(m)>=length(GT_sfc) GT_operation(m)=length(GTsfc); end FC-operation(m)=fix(100*FCloading(m)/(n*FCpowerrange)); if FC-operation (m) ==o FC-operation(m)=1; end if FC-operation(m)>=length(FCsfc) FC-operation(m)=length(FC_sfc); end FC efficiency(m)=FC sfc(FC-operation(m)); GTefficiency(m)=GT sfc(GT operation(m)); %Find Combined fuel consumption and compare it to GT only value: Potentialfuel consumption(1,n)=FC loading(m)*FCefficiency(m)+GTsfullyloaded*GTpower-range*GT _sfc(length(GTsfc))+Partial_GTload(m)*GTefficiency(m); if Fuel consumption(l,n)>=Potentialfuelconsumption(l,n) Fuelconsumption(l,n)=Potentialfuelconsumption(l,n); Best use-ratio(l,n)=m/100; end end end Fuelconsumption differential (l,n)=GTonlyconsumption(l,n) -Fuelconsumption(l,n); end Cumulativefuelconsumption(1,n)=Fuel_consumptiondifferential(1,n); %establish fixed point to integrate from for a=2:(length(Electric required)-1) Cumulativefuelconsumption (a,n) =Cumulativefuel consumption (a1,n)+Fuelconsumption differential(a-l,n); end Cumulativefuel consumption (length (Electricrequired), n) =Cumulative-fuel-consumption (length (Elect ric_required)-1,n)+Fuel consumptiondifferential(length(Electric required),n); end %GTonlyconsumption=GTonlyconsumption*grams_tolbs*lbstoLT; %Fuel-consumption=Fuel_consumption*grams to lb*lbsto_LT; %LT Fuelper_1000ONM=sum(GT only consumption) -sum(Fuel consumption); %Input Ship Profile Data: Power array = [5 10 15 20 25 30 35]; % Establish knots Power conversion Array (in 5 knot increments) Power required = [500 1600 5200 12000 27000 66500 100000]; %KW resistance to overcome Endurance = 4000; %Nautical miles endurance Speedprofile = [.119 0 .466 .356 .044 .015 0]; %Profile from ASSET 5.3 Flight 1 Mission Analysis Module (.11305 %Speed profile %Speed profile %Speedprofile %Speedprofile %Speed profile 0 .4427 ,366483333 .0418 .025483333 .010483333); [.1073975 0 420565 .3764425 .03971 .0354425 .0204425); [.102027625 0 .39953675 .385903708 0377245 .044903708 .029903708); [.096926244 0 .379559913 .394891856 .035838275 .053891856 .038891856); [.092079932 0 .360581917 .403430597 .034046361 .062430597 .047430597); %Speedprofile %Speedprofile %Speedprofile %Speedprofile %Speed profile [0.119 [0.119 [0.119 [0.119 [0.119 %Speedprofile = [1/7 0 0 0 0 0 0.4194 0.3728 0.3262 0.371533333 0.044 0.2796 0.418133333 0.044 0.433666667 0.044 0.233 0.030533333 0.015533333); 0.387066667 0.044 0.046066667 0.031066667); 0.4026 0.044 0 0616 0.0466]; 0.077133333 0.062133333]; 0.092666667 0.077666667]; 1/7 1/7 1/7 1/7 1/7 1/7] ; %Average use distribution _Seed profile Speed profile %Speed profile %Speed profile %Speed__profile %Speed profile %Speed profile .25 025 =. =C 0 = I/5 = = 113 0 iNormal 1ist=ibuti3r. 0251 075 2 .2 A High speed ops .25 .51; 12 025 .05 Low speed cps 25 .2 1 05 .025 025 0 0 1 0 0; iTest profile 11/5 1/5 0 1/5 0 1/'51 ; %No pickup effect 1/3 0 0 1/3 0 £1; %least preferred piofale (high argument 0 1/3 0 1/3 1/31; %optimal GI profile 075 for fuel cells 0 0 0 0 0 0); 5 knots dominant %r0 knots dominant 1 0 0 7 0 0 01; [ 2Speed5profile knots.15dominant . 0 0 0 01; 1 [0 0 0 0 1 0 0 0, %20 knots dominant =C [0 0 0 0 1 0 01; %25 knots dominant 1 0 0 0 0 1 01; %30 knots dominant 0 0 0 0 0 0 !1; %30 knots dominant %Speed profile = Speedp0of2le Speed3profile %Speed profile %speed profile %Speed profile normalize3=GTparallel_consumption all (length(GTparallel_consumption all)); figure; hold on; plot (10*Electric required/length (Electric_required), GT onlyconsumption/normalize3) %plot (Electric required/length(Electric required),GTparallel consumption two/normalize3) plot(10*Electricrequired/length(Electric_required),GTparallel consumptionall/normalize3) xlabel('Power Range (MW)') ylabel('GT Fuel Consumption (Ratioed to 100 MW amount)') figure; hold on; plot(10*Electricrequired/length(Electric_required), (GT-parallel consumptionallGTonlyconsumption)/ (normalize3)) xlabel ('Power Range (MW)') ylabel('Parallel to Series GT Operation Fuel Consumption Differential') normalize2=GT onlyconsumption (length(GT_onlyconsumption)); figure; hold on; plot(10*Electricrequired/length(Electric_required),GT-onlyconsumption/normalize2) plot(10*Electricrequired/length(Electric_required),Fuelconsumption/normalize2) xlabel('Power Range (MW)') ylabel('Fuel Consumption (Ratioed to rate at 100 MW)') figure; hold on; plot(Electric_required/10A3,Fuel_consumption differential/(normalize2)) %plot (Electric_required,Cumulative fuel consumption/75) figure; hold on; plot(Power required/1000,Powerarray) figure; hold on; plot((l:length(Bestuseratio))/100,Bestuseratio); if length(Power-array) -= length(Speedprofile) ;%confirm speeds used alligns with powers available display ('Speed Profile is incorrect'); end %Speedsused = length(Speedprofile) ; %Set number of speeds in the profile %Profilecheck = sum(Speed profile); %Verify profile is complete %if Profile check -= 1.0000; display ('Speed Profile is incorrect'); % %end %Total fuel saved test(n)=0; Total fuel saved(FC available)=0; TotalfuelusedGTonly(FCavailable) =0; Total fuelusedhybrid (FC available) =0; for n=1:FC available for k=l:length(Speedprofile) Power for speed(k)=fix(length(Electricrequired)*Power required(k)/Electric required(length(Elect ric_required))); if Powerfor speed(k)<=1 Powerfor speed(k)=1; end if Power for speed(k)>=length(Electric required) Power forspeed(k)=length(Electricrequired); end FuelusedGTonly (k,n) =Endurance*Speedprofile (k)*GT onlyconsumption (Power for speed (k),n); Fuelused_hybrid (k,n) =Endurance*Speedprofile (k)*Fuelconsumption (Power for-speed (k),n); Fuelsaved (k,n) =Endurance*Speed_profile (k)*Fuelconsumption differential (Power_forspeed (k),n); TotalfuelusedGTonly(n)=TotalfuelusedGTonly(n)+FuelusedGTonly(k,n); Totalfuel used hybrid (n)=Totalfuelusedhybrid (n)+Fuel_usedhybrid (k,n); Totalfuelsaved(n)=Totalfuelsaved(n)+Fuelsaved(k,n); %Total fuel saved test (n)=Total fuel saved test (n)4Fuel usedGT only(k n) Fuelusedhybrid(k,n); %Test=Total fuel saved(n)-Total fuel saved test(n) end end %Total fuel usedGT_only(n) %Total fuelused hybrid(n) %for n=l:FC available % Total fuel saved(n) %end TFUplotl = [0.9904 0.9778 0.9779 0.9032 0.8865 0.8767 0.8745 0.8744 0.8805]; TFUplot2 = [0.9918 0.9854 0.9782 0.8762 0.8784 0.9811 0.9175 0.9811 0.9033 0.9814 0.8945 0.9810 0.8936 0.9791 0.8930 0.9727 0.8935 0.8949 0.8931 0.8964 0.8959 0.8972 0.8929 0.8930 0.8982]; TFUplot3 = [0.9929 0.9836 0.9287 0.9837 0.9164 0.9875 0.9892 0.9845 0.9777 0.8750 0.8777 0.9868 0.9886 0.9832 0.9755 0.8743 0.8793 0.9857 0.9876 0.9821 0.9679 0.8750 0.8743 0.9847 0.9868 0.9808 0.9590 0.8739 0.8756 0.9837 0.9651 0.8926 0.8940 0.9859 0.9839 0.9087 0.9836 0.9079 0.9819 0.9074 0.9764 0.9079 0.9091 0.9075 0.9074 0.9120]; TFUplot4 = [0.9938 0.9857 0..9857 0.9375 0.9268 0.9204 0.9190 0.9189 0.9229]; TFUplot5 = [0.9945 0.9104 0.9099 0.9111 0.9074 0.9083 0.9906 0.9859 0.9201 0.9215 0.9900 0.9856 0.9194 0.9211 0.9892 0.9842 0.9189 0.9221 0.9884 0.9793 0.9194 0.9189 0.9876 0.9736 0.9186 0.9197 0.9873 0.9874 0.9876 0.9873 0.9448 0.9352 0.9293 0.9287 0.9283 0.9287 0.9296 0.9283 0.9284 0.9318]; 0.9306 0.9302 0.9311 0.9283 TFUplot6 0.9783 0.9115 [0.9909 0.9784 0.8966 0.9858 0.9787 0.8873 0.9850 0.9782 0.8864 0.9837 0.9762 0.8857 0.9825 0.9694 0.8863 0.9916 0.9911 0.9904 0.9860 0.9898 0.9817 0.9698 0.9071 0.9891 0.9792 0.9476 0.8772 0.8776 0.9822 0.9554 0.8954 0.8958 0.9846 0.9614 0.9095 0.9777 0.9348 0.8772 0.8740 0.9208 0.8735 0.8771 0.9810 0.9445 0.8954 0.9325 0.8922 0.8926 0.8953 0.9835 0.9520 0.9095 0.9416 0.9068 0.9098 0.9071 0.9095 0.9866 0.9662 0.9207 0.9210 0.9856 0.9580 0.9208 0.9187 0.9489 0.9184 0.9207 0.9881 0.9873 0.9766 0.9701 0.9628 0.9548 0.9280 0.9299 0.9299 0.9278 0.9281 0.9299 0.9782 0.9399 0.8883 0.9273 0.8849 0.9290 0.9813 0.9615 0.8853 0.9302 0.9796 0.9513 0.8883 .. . ... 0.8878 0.8859 0.89141; 0.8857 TFUplot7 = [0.9925 0.9818 0.9818 0.9197 0.9310 0.9116 0.9131 0.9158]; 0.9114 TFUplot8 = [0.9938 0.9844 0.9844 0.9370 0.9456 0.9320 0.9308 0.9306 0.93411; TFUplot9 = [0.9947 0.9864 0.9864 0.9504 0.9569 0.9457 0.9466 0.9483]; 0.9454 TFUplot10 = [0.9955 0.9880 0.9880 0.9611 0.9659 0.9575 0.9582 0.9595]; 0.9573 plot11 = 0.9710 0.8386 0.8234 0.8310]; plot12 = 0.9723 0.8514 0.8366 0.8441]; plot13 = 0.9746 0.8766 0.8627 0.8697]; plot14 = 0.9758 0.8892 0.8757 0.8825]; 0.8857 0.8869 0.8887 0.8854 0.9114 0.9854 0.9750 0.9119 0.9150 0.9115 0.9844 0.9690 0.9112 0.9124 0.9829 0.9612 0.9134 0.9137 0.9817 0.9526 0.9134 0.9112 0.9894 0.9843 0.9311 0.9323 0.9884 0.9831 0.9307 0.9334 0.9876 0.9792 0.9309 0.9308 0.9866 0.9745 0.9305 0.9314 0.9853 0.9686 0.9321 0.9324 0.9843 0.9621 0.9321 0.9305 0.9548 0.9301 0.9321 0.9908 0.9863 0.9459 0.9468 0.9900 0.9854 0.9455 0.9477 0.9892 0.9824 0.9457 0.9457 0.9884 0.9789 0.9455 0.9461 0.9872 0.9744 0.9467 0.9469 0.9863 0.9694 0.9466 0.9454 0.9639 0.9451 0.9466 0.8893 0.8887 0.9882 0.9821 0.9127 0.9142 0.9875 0.9817 0.9120 0.9137 0.9900 0.9846 0.9316 0.9327 0.9914 0.9866 0.9462 0.9471 0.8902 0.9864 0.9801 0.9887 0.9920 0.9925 0.9912 0.9905 0.9898 0.9872 0.9850 0.9823 0.9790 0.9881 0.9879 0.9574 0.9577 0.9579 0.9575 0.9574 0.9583 0.9584 0.9583 0.9591 0.9576 0.9579 0.9586 0.8882 0.9430 0.9108 0.9133 0.9879 0.9753 0.9582 0.9573 0.9712 0.9570 0.9582 0.9709 0.8597 0.8261 0.8233 [0.9870 0.9710 0.8255 0.8283 0.9805 0.9623 0.8241 0.8275 0.9794 0.9538 0.8232 0.8295 0.9777 0.9762 0.9314 0.8226 0.8248 0.9747 0.9165 0.8268 0.8274 0.9726 0.9436 0.8240 0.8231 0.8268 0.8227 0.9707 0.8820 0.8221 0.8268 [0.9879 0.9728 0.8387 0.8414 0.9816 0.9703 0.8373 0.8406 0.9805 0.9623 0.8364 0.8425 0.9789 0.9527 0.8373 0.8364 0.9775 0.9412 0.8359 0.8380 0.9760 0.9270 0.8399 0.8405 0.9739 0.9111 0.8400 0.8360 0.9721 0.8937 0.8354 0.8399 0.9723 0.8720 0.8393 0.8365 (0.9892 0.9750 0.8646 0.8672 0.9833 0.9744 0.8633 0.8664 0.9823 0.9747 0.8625 0.8683 0.9809 0.9705 0.8633 0.9795 0.9605 0.8619 0.8639 0.9781 0.9475 0.8658 0.8663 0.9761 0.9328 0.8658 0.8620 0.9744 0.9165 0.8615 0.8657 0.9746 0.8960 0.8652 0.8625 [0.9898 0.9762 0.8776 0.8800 0.9842 0.9756 0.9832 0.9758 0.8755 0.8811 0.9818 0.9755 0.8763 0.8755 0.9805 0.9792 0.9577 0.8787 0.8792 0.9772 0.9436 0.8788 0.8751 0.9756 0.9279 0.8745 0.8787 0.9757 0.9080 0.8781 0.8756 0.8763 0.8793 0.8624 0.9700 0.8750 0.8769 normalize=TotalfuelusedGT-only(1); figure hold on; ylim([0.8 1]) %plot (l:FC availableTotal fuel saved/normalize) plot (1: FCavailable,Totalfuel_usedhybrid/normalize) plot(1:FCavailable,TotalfuelusedGT only/normalize) %plot (1:FC availableplotll) %plot(1:FC available,plotl2) %plot (1:FCavailable,plot13) %plot (1:FC_available,plot14) %plot(l:FCavailableTFUplotlo) ylabel('Total Fuel Consumption (ratioed to GT only plant)') xlabel('Number of Installed Fuel Cells') 0.9001